The parents of 746 children enrolled in grades from kindergarten through high school in 1985 were interviewed in a national study designed to identify factors affecting parent involvement in schools and education-related activities. Three kinds of parent involvement were measured: (1) willingness to visit the child's school three or more times during a year, or personalistic involvement; (2) active participation in a parent-teacher association, or supportive involvement; and (3) attentiveness to local school issues, or policy-oriented involvement. A logit analysis was conducted for each of these three types of involvement to develop three models of parental participation. The models revealed that the parent's gender was the strongest predictor of participation in all three cases, that the parent's educational level was a strong predictor of policy-oriented involvement, and that the child's grade level was a strong predictor of personalistic and supportive involvement. The demands placed on parents' time did not appear to affect the impact of gender on participation. Parents of younger children proved more child-oriented; this factor affecting participation appeared unrelated to gender or to parents' educational levels. The study suggests that parental involvement is a multidimensional construct that deserves further examination. (PGD)